PROJECT SUMMARY/ABSTRACT There is a fundamental gap in the understanding of why Alzheimer's disease (AD), which is defined by amyloid plaques and tau tangles, exhibits such a diverse topographic distribution of neuropathology and range of clinical variability. To operationally classify the heterogeneity of AD, the PI designed a mathematical algorithm using differences in hippocampal and cortical tangle patterns. From these studies, three AD variants were identified ? hippocampal sparing (HpSp), typical, and limbic. Striking clinical differences were revealed, as exemplified by an aggressive clinical course identified in HpSp AD cases compared to typical AD; and in direct contrast to limbic AD. As such, we propose an extension to the theoretical concept of heterogeneity in AD by drawing attention to and investigating limbic AD as an insidious form of the disease. Additionally, HpSp and limbic variants of AD are relatively under-appreciated in current dementia care, thus a better understanding of these entities is a genuine public health imperative. The overall objectives are to leverage one of the largest autopsy-confirmed AD series to decipher neuropathologic features, investigate clinical heterogeneity, and utilize longitudinal neuroimaging to differentiate these AD variants. Our central hypothesis is that hippocampal- involving and hippocampal-sparing AD share disease traits, but demographics, phenotypic presentation, and longitudinal neuroimaging can be used to assess rate of disease progression because they significantly differ between HpSp and limbic AD variants. In Aim 1, novel neurobiologic insights into AD will be revealed by investigating differential neuropathologic changes of altered proteins, neuronal loss, and key subcortical nuclei AD. This will be performed using innovative digital pathology to quantify neuroanatomic distribution of pathologic severity. In Aim 2, contributors to variability in rate of disease progression across AD variants will be identified by examining heterogeneity of demographics and phenotypic presentations. To accomplish this aim, an in-depth investigation of each case's clinical history and functional decline will be performed. Finally in Aim 3, incorporation of longitudinally-collected antemortem neuroimaging characteristics of autopsied AD cases will enable dynamic interpretation of selective hippocampal and cortical vulnerability. To accomplish this aim, magnetic resonance imaging will be investigated using sophisticated software to assess atrophy measures. The contribution of this proposal will be significant because it will establish an objective antemortem approach to classifying heterogeneity in AD by integrating clinical variability and neuroimaging patterns using a well- established prospectively followed cohort who have been autopsied. The proposed research is innovative because it will evaluate one of the largest cohorts of autopsy-confirmed AD cases, objectively classify AD variants using neuropathology as the gold-standard, evaluate neuropathologic associations with longitudinal measures of memory and functional decline, employ automated methods to quantify brain pathology, and use state-of-the-art neuroimaging methods to recapitulate the AD variant classification algorithm.